PROJECT SUMMARY Depressive disorders (DD) are one of the most prevalent forms of mental disorders, and are associated with significant mortality, morbidity, and economic costs. The incidence of DD begins to increase in early-mid adolescence and peaks in emerging adulthood. The Stony Brook Temperament Study is a longitudinal study that seeks to identify early behavioral and neural precursors and risk factors for DD and understand the neurobiological and psychosocial processes through which these early manifestations develop into clinically significant psychopathology. We initially assessed a large community sample of 3-year old children, and have followed them at 3-year intervals through age 15. This competing renewal application seeks to re-evaluate the sample at ages 18 and 21. Specifically, we propose to trace pathways from laboratory observations of temperamental low positive emotionality and high negative emotionality at age 3 to neural indices of emotional and social information processing and DD symptoms and onsets from middle childhood through emerging adulthood. In addition, we seek to understand how life stress over the lifespan and the developmental challenges of emerging adulthood contribute to risk for DD, and to examine the specificity of these risk pathways in comparison with anxiety and substance use disorders. Finally, we will use cutting-edge machine learning techniques to maximize prediction of DD and understand which domains and developmental periods contribute most to prediction. This information about risk pathways and processes should contribute to understanding when and how to intervene in order to prevent DD and limit their progression.